Comparative Evaluation of Machine Learning Algorithms for Heart Disease Prediction Using the UCI Cleveland Dataset

This study presents a comparative evaluation of three supervised machine-learning algorithms—Logistic Regression, Random Forest and Support Vector Machine (SVM)—for heart disease classification using the Cleveland subset of the UCI Heart Disease dataset. The dataset contains 303 records and 13 predictor variables. The original target variable was transformed into a binary classification outcome, distinguishing observations with no recorded heart disease from those with disease present. The models were evaluated using an 80:20 stratified train-test split and five-fold stratified cross-validation. Performance was assessed using accuracy, precision, recall, F1-score and receiver operating characteristic area under the curve (ROC-AUC). On the held-out test set, Random Forest achieved the highest accuracy of 90.16%, while Logistic Regression achieved the highest ROC-AUC of 95.78%. Cross-validation showed a different performance pattern, with Logistic Regression achieving the highest mean accuracy of 84.81%. The findings demonstrate that model performance varies across evaluation methods and metrics. Given the small, historical dataset and the absence of independent external validation, the models should be considered research models rather than clinically validated diagnostic tools.